Feb 21, 2014 (7 years and 10 months ago)


Clays and Clay Minerals, Vol. 40, No. 5, 586-592, 1992.
Department of Mining, Metallurgical and Petroleum Engineering
University of Alberta, Edmonton, Alberta, Canada T6G 2G6
Abstract- The sedimentation behavior of a fine kaolinite, comprising a substantial proportion of colloidal
particles as well as non-colloidal ones, has been studied when fresh FeC13 or F2(SO4)3 electrolytes are
added. The sedimentation behavior depends on the pH and the nature of electrolytes and can be explained
qualitatively, in our study, by the theory of Derjaguin, Landau, Verwey, and Overbeek (DLVO theory).
Fe helps also to aggregate the kaolinite particles in flocs. Two extreme kinds of qualitative sedimentation
have been observed: flocculation-sedimentation and accumulation-sedimentation. However, the transi-
tion between the two kinds of sedimentation is quite progressive. The present results are discussed in
reference to the DLVO theory and the hydrolysis behavior of Fe electrolytes.
Key Words--Colloid, Flocculation, Iron, Kaolinite, Sedimentation.
The behavior of raw materials in aqueous media
plays an important role in ceramic processing. It influ-
ences the formation of crack-free wet parts from the
powder, as well as the sintering and the quality of a
final product. Because it is a major raw material for
ceramics, the flocculation and dispersion behavior of
kaolinite in aqueous media has been studied exten-
When an electrolyte is added, the flocculation of a
kaolinite suspension depends on the valence and the
nature of the ions, ionic strength, and pH (Hiemenz,
1977; van Olphen, 1977; Rand et al., 1977; Bolland et
al., 1976; Arora et aL, 1979; Swartzen-Allen et al.,
1976; Goldberg et al., 1987). The stability of a colloidal
system can often be explained by the DLVO theory
(Hiemenz, 1977). When the electrical repulsion of
identically charged double layers around particles is
strong enough, the colloidal system remains dispersed
(Swartzen-Allen et al., 1974). Flocculation is brought
about by the decrease of this repulsion. However, in
clay suspensions the situation is complicated by dif-
ferent charges on the edges and faces of platelike par-
ticles (Swartzen-Allen et al., 1974; van Olphen, 1977;
Worrall, 1986). For kaolinite, the reported values for
the point of zero charge (z.p.c.) are between pH 5.5
and 6.7, depending on the conditions (Young et al.,
1987). Below the z.p.c, the edges of kaolinite particles
are considered to be charged positively and the faces
are charged negatively. Above the z.p.c, the edges also
become negatively charged, and thereby increase the
stability of the clay dispersion (Goldberg et al., 1987;
Swartzen-Allen et al., 1976; Bolland et aL, 1976). Con-
sequently, for kaolinite, a number of interaction modes
has been proposed which call for face-to-face (FF), edge-
Copyright  1992, The Clay Minerals Society
to-face (EF) or edge-to-edge (EE) particle associations
(van Olphen, 1977; Rand et al., 1977; Flegmann et al.,
It is desirable to elucidate the actual structure of clay
aggregates. For this purpose, a technique was success-
fully applied by Zou and Pierre to mont mori l l oni t e
(1992). Relatively strong montmorillonite flocs, with
respect to mechanical compaction, were made with
fresh Fe 3+ electrolyte and the flocs were dried by the
supercritical method. Fe 3§ is known to play a triple role
on the flocculation-dispersion behavior of clay sus-
pensions: (l) as a counterion; (2) as an exchange cation
anchoring Fe hydrolysis complexes to the surface of
clay; (3) as a bonding agent between clay particles by
the intermediate of the Fe hydrolysis complexes
(Greenland, 1975; Young et al., 1987; Rengasamy et
al., 1977; Goldberg et al., 1987).
In a preliminary step, it is necessary to determine
the pH and Fe electrolyte concentration levels where
flocculation occurs with kaolinite. This was the purpose
of the present work. Aqueous kaolinite suspensions
with a particle size ranging from colloidal to non-col-
loidal, according to the supplier, have been studied.
Their sedimentation kinetics in the presence of Fe elec-
trolytes are reported and discussed. The macroscopic
aspects of the sediments, with respect to thickness and
qualitative fluidity are described.
The kaolinite studied was Hydrite UF from the
Georgia Kaolin Company, Inc. The median particle
size was 0.20 ~m and the mass proportion of colloidal
particles with size - 0.20 ~m was about 55%, according
to the supplier (Georgia Kaolin Company, 1990). The
specific area determined by the BET was 20.1 mVg,
Vol. 40, No. 5, 1992 Sedimentation of kaolinite 587
which is consistent with the value provided by the
This kaolinite was treated according to the met hod
of Schofield and Samson (1954) with 1M NaC1 solution
adjusted at pH 3 with HC1 to convert the kaolinite to
the sodium form. The treated kaolinite was washed
with distilled water until there was no evidence of CI
in the supernatant liquid tested by AgNO3. Treated
kaolinite (0.5 g) was mixed with 50 g distilled water,
then 1 ml of 0.5 N Na4P207 solution was added to
disperse the kaolinite. The kaolinite suspension pH was
9.30 _+ 0.05. This process insured that the clay particles
carry only negative charges both on edges and faces,
as the phosphate anions are not easily desorbed. This
should permit the comparison of the results with the
predictions from the DLVO for platelike particles
(Pierre, 1992), after a further structure study by su-
percritical drying is achieved.
In the present study, the colloidal interactions were
modified by changing the pH, and/or adding a fresh
Fe 3+ electrolyte. The electrolytes we studied most ex-
tensively were fresh
and fresh
Fe2(SO4) 3.
In order to avoid interaction between different an-
ions, the pH values of the kaolinite suspensions were
adjusted with NaOH to 10.0 and 11.0, and with HCI
or n2s o 4 t o 8.0,
6.0, 4.0, and 2.0, depending on the
anion in the electrolyte. For example, HCI was used if
the effect of FeC13 was investigated, while HzSO4 was
used with
Fe2( SO4) 3.
At each pH, six different FeC13 or
Fe2(SO4)3 concentrations were examined: 0.0, 0.5, 1.0,
2.0, 5.0, and 10.0 mN (meq/liter). Fresh
FeC13, or
Fe2(SO4)3, was always prepared just prior to an exper-
Each suspension was prepared in a graduated cyl-
inder with an inside diameter of 28 mm, which was
large enough to neglect the wall effect (Michaels
et al.,
1962). Then, distilled water and a proper amount of
fresh 20 mN electrolyte solution were added to get an
initial suspension vol ume of 100 ml. The cylinders
were turned upside down 20 times to uniformly dis-
tribute the component s inside the suspensions. All ex-
periments were performed at room temperature and
the cylinders were covered with a piece of parafilm to
prevent water evaporation.
To investigate the role of different cations and anions
on the sedimentation behavior, supplementary exper-
inaents were carried out with the electrolytes FeSO4,
NaCI and Na2SO 4 at concentrations of 0.5, 1, 2, 5, and
10 mN at pH 2.
In all sedimentation experiments where a clear in-
terface occurred, the position of this interface was read
every 5 or l 0 minutes during the first hour, then every
hour for the following two hours, and finally every 24
hours until 360 hours. The rate at which this interface
moved was used to measure the accumulation rate, or
the settling rate (depending on sample behavior), in
Figure 1. Sedimentation of 0.5% Na-kaolinite suspensions
at pH = 4.0 after 3 day settling: (a) accumulation-sedimen-
tation with
FeCl 3 =
10 raN; (b) flocculation-sedimentation
wi t h FeE13 =
1.0 mN; (c) mixed accumulation-flocculation
sedimentation with FeC13 = 5.0 mN.
The sedimentation behavior of kaolinite suspensions
depended largely on the electrolyte concentration, the
cation and the anion in the electrolyte, and the pH.
Three types of sedimentation behavior were observed.
Two of these, called flocculation-sedimentation and
accumulation-sedimentation, were observed earlier for
montmorillonite (Zou and Pierre, 1992). However, they
were called, respectively, two-layer and three-layer be-
havior in the previous publication. The third sedi-
mentation behavior is called mixed flocculation-ac-
cumulation here. The three sedimentation behaviors
can be described as follows.
Figure 1 a shows accumulation-sedimentation. From
the bottom to the top of the test cylinder one sees
successively: (1) the accumulated sediment; (2) the re-
maining diffuse suspension of particles; (3) the clear
supernatant liquid. A sharp interface separates the ac-
cumulated sediment and the remaining suspension. This
interface kept movi ng up with time from the beginning
to the end of the experiment. That is to say, the ac-
cumulated sediment thickness increased with time. Also
this accumulated sediment did not flow easily when
tilting the cylinder. In contrast, the transition zone be-
tween the remaining suspension and the supernatant
liquid is not sharp. With a spectrophotometer, it should
be possible to define arbitrarily an interface corre-
sponding to a given turbidity of the liquid, but we did
not have the instrumentation to do so. However, qual-
itatively, the remaining suspension kept settling down
by feeding the accumulated sediment, until it disap-
peared completely.
Figure l b shows flocculation-sedimentation. A sin-
gle sediment layer is separated by a sharp interface with
the clear supernatant liquid. This interface kept mov-
ing down from the beginning to the end of sedimen-
tation. That is to say the sediment kept shrinking until
58 8 Ma and Pierre Clays and Clay Minerals
t -
g 1o
~" 0
FeCI 3 concentration
# 1.0 mN
m 2.0 mN
I ni t i al seal i ng r at e
1 I i I
100 200 300 400
500 600
i1.o .
~ 0.5
~00 t
. - =- -" ,
< 0 100
FeCI 3 concentration
= O.OmN
 lO.OmN
I I I I I 8
200 300 400
I t
FeCI 3 concentration
'~ 0.5raN
 5.0raN
Accumulated sedi ment
| I |
100 200 300 400 500 600
Time (min.)
Figure 2. Displacement rate of sharp interfaces in 0.5% Na-
kaolinite suspensions at pH = 4.0: (a) flocculate-supernatant
liquid interface with FeCI3 concentrations 1.0 and 2.0 mN;
Co) accumulated sediment-remaining suspension interface with
0.0 and 10.0 mN of FeC13; (c) flocculate-supematant liquid
and accumulated sediment-remaining suspension with 0.5 and
5.0 mN of FeCl3.
it reached a final volume. Also, this sediment would
flow easily when tilting the cylinder.
Mixed sedimentation behavior begins as an accu-
mulation-sedimentation in Figure la, with a sharp in-
terface between the accumulated sediment and the re-
mai ni ng suspension. This sharp interface moves up.
With time, however, the diffuse suspension slowly floc-
culates. In spite of the fact that its interface with the
bottom sediment remained sharp (well-defined line),
the contrast in turbidity between each side of the in-
terface slowly decreased. After some time, depending
on the visual acuteness of the experimentalist, it was
no longer possible to observe this sharp interface. Fig-
ure 1 c shows the stage where this sharp interface is no
longer visible. Only one sediment layer is visible and
its transition-zone with the clear supernatant liquid is
diffuse. However, after a longer time, this transition-
zone transformed from a diffuse-zone to a sharp in-
terface. The end of the sedimentation experiment looked
like flocculation-sedimentation as in Figure lb, with
apparently one single sediment separated from the top
supernatant liquid by a sharp interface. This sharp in-
terface kept movi ng down.
Flocculation-sedimentation kinetics are reported for
FeCI3 concentrations of 1.0 and 2.0 mN at pH 4.0 in
Figure 2a. It was characterized, at the beginning, by an
initial period when no obvious settling occurred, on
the order of 10 minutes. During this induction period,
flocs were being formed. At some moment, these flocs
joined to each other and created an apparently uniform
sediment, separated by a sharp interface from the clear
supernatant liquid. Settling of the flocculated sediment
started at a relatively fast initial settling rate. Later, the
settling rate began to slow down progressively, until it
reached an aging stage when the sediment thickness
decreased very slowly, a process known as syneresis in
sol-gel science.
The sedimentation data with FeC13 concentrations
0.0 and 10.0 mN at pH 4.0 are reported in Figure 2b.
They are of the accumulation-sedimentation type. This
bottom accumulated sediment was formed by the fas-
test-settling clay particles, the largest and heaviest clay
particles. Its initial volume was zero and it increased
slowly by accumulation of new particles. On top of the
accumulated layer, the remaining suspension was com-
prised of smaller, slower-settling clay particles. They
constituted a colloidal sol, showing that no extensive
flocculation had occurred. In agreement with the in-
formation by the supplier, that the particles were not
of uniform size, the interface between this colloidal sol
and the supernatant liquid was not sharp, but diffuse.
The final sediment volume in accumulation-sedimen-
tation was much lower than that in flocculation-sedi-
mentation. Therefore, the accumulated sediment was
more densely packed than the flocculated sediment.
Also, no syneresis occurred.
In this study, the mixed flocculation-accumulation
sedimentation behavior occurred in conditions inter-
mediate between those for flocculation- and accumu-
lation-sedimentation. The evolution of the sediment
volumes is reported in Figure 2c for FeCI 3 concentra-
tions of 0.5 mN and 5.0 raN. The data show initially
the accumulation rate of the bottom sediment as long
as it could be followed. They also show, in the end,
the settling-rate of the flocculated-sediment on top of
the accumulated sediment, starting from the moment
when a sharp interface with the supernatant liquid could
be observed. For 5 mN FeC13, the data points for these
2 periods form a single curve with a maxi mum in the
sediment volume after approximately 100 min. How-
Vol. 40, No. 5, 1992 Sedimentation of kaolinite 589
-O O    "~'0
.-o     Fi occul at ed/o '
  o   
Accumul at ed
4 0   ,."  ",, 
  %%
- / %
 %
/ %
.4  0// 0 0 x 
/' %
Mi xed
t / "" \
5, /o -
0  0 ~ 
Flocculated ~ ~
t 0 0
f 0
! 0
0 0
0 0
11.0~   
IO.0~.+   
90~e  
' / Accumulated
8.0,-   t
7.0 /"
6.0  O/O
5.0 ~ ,,z Mixed
z0t go t~ ? I
0.0 1.0 2.0 3.0
Fl occul at ed
0 0
0,0 1.0 2.0 3.0 4,0 5.0 6.0 7.0 8.0 9.0 I 0.0 4.0 5.0 6.0 7.0 8.0 9,0 I 0.0
Fe CI s ( r aN) Fe 2( SO4 ) 3 (raN)
(a) (b)
Figure 3. Diagrams of the sedimentation behavior in 0.5% Na-kaolinite suspension with: (a) FeCI3; (b) Fe2( 504) 3. The black,
white, and mixed dots indicate accumulation, flocculation, and mixed behavior, respectively.
ever the points do not concern the same interface. Re-
cent SEM observations show that a thin layer of floc-
culated sediment is on top of the bottom accumulated
sediment (Ma and Pierre, 1992). This is substantiated
by the data for 0.5 mN FeC13, where the evolution of
the accumulated sediment volume could only be fol-
lowed for a short time. This interface did not disappear
on reaching the top of the remaining suspension. It
disappeared by progressive loss of contrast between the
accumulated and the flocculated sediment. Actually,
this mixed sedimentation behavior occurred when floc-
culation was slow. The biggest clay particles settled
without participating significantly in flocculation, while
the smallest particles remained in suspension long
enough to achieve substantial flocculation.
Diagrams summarizing the sedimentation behavior
of the kaolinite suspensions are shown in Figure 3. The
dotted lines have been tentatively drawn to outline the
ranges of conditions where each type of sedimentation
occurred. The sedimentation behavior not only de-
pended on pH and the Ire 3§ electrolyte concentration,
but also on the nature of anions. With Fe2(SO4)3, ac-
cumul at i on-sedi ment at i on occurred in a smaller do-
mai n of conditions than with FeC13, and the mixed
sedimentation regime occurred in only two experi-
ments. By comparison, the range of conditions where
flocculation-sedimentation occurred was greatly ex-
With sedimentation data similar to Figure 2a (which
corresponds to what is called flocculation-sedimenta-
tion here), previous researchers defined what they called
a constant settling rate. This rate describes the begin-
ning of sedimentation when the sediment volume de-
creases linearly with time (Michaels et al., 1962). This
constant settling rate lasted as long as 30 minutes with
kaolinite and is called the initial settling rate in the
present study (Figure 2a). We also defined, in the pres-
ent study, the final sediment volume as the sediment
volume after 360 hours' settling. Figure 4 and Figure
5 show the initial settling rate and final sediment vol-
ume, respectively, for increasing concentrations of dif-
ferent electrolytes at pH = 2.0. By comparing the data
for FeC13 and Fe2(SO4)3, it appears that the nature of
the ani on had a big influence. By comparing the data
for NaCl and FeC13, it appears that the cation also had
a big influence. Hence, the combi nat i on of the ani on
and cation is important.
The data for initial settling rate and final sedimen-
tation volume are reported in more detail for Fe2(SO4)3,
in Figure 6 and Figure 7, since flocculation-sedimen-
ration occurred in a large range of conditions with this
electrolyte, as indicated in Figure 3. In Figure 7, the
"-~ 35
~ 30
e~ 25
15 Fe2(SO4) 3
5 . . I . . i . . , . . , , r i i t , , : , , i . . i . ,
0 1 2 3 4 5 6 7 8 9 10
Cation Concentration (mN)
Figure 4. Initial settling rate of 0.5% Na-kaolinite suspen-
sion flocculated by different electrolytes at pH 2.0.
590 Ma and Pierre Clays and Clay Minerals
3.5 FcCI 3
~ 3.0
~-" 2.5
.~_ ~
' ~ NaC1
N ~ ~ ~ ~ NaCI
L.ol . , i , . I , , v , , ,   ,  . , . . i  - ,   , . .
0 1 2 3 4 5 6 7 8 9 10
Cat i on Concent r at i on ( r aN)
Figure 5. Final sediment thickness, after 360 hours sedi-
mentation, of 0.5% Na-kaolinite suspension flocculated by
different electrolytes at pH 2.0.
Fl occul at ed sedi ment s
~ 2
 pH = 6.0
.~ .Or .- 0 pH = 9.3
~'~ " pH = 11.0
.k.4 
0 " " ' ' " ' ' " ' " " ' " " ' " " ' " ' ' " " ' " " ' " "
1 2 3 4 5 6 7 8 9 10
Fe2 ( SO4) 3 Concent r at i on ( r aN)
Figure 7. Final sediment thickness, after 360 hours sedi-
mentation, of 0.5% Na-kaolinite suspensions as a function of
Fe=,(SO4)3 concentration.
final sediment vol umes when accumulation-sedimen-
tation occurred are also reported. They correspond to
the dotted lines. A small initial amount of the electro-
lyte drastically increased the final sediment volume.
However, a further increase in electrolyte concentra-
tion sightly decreased the final volume.
By comparing Figure 6 with Figure 7, it appears that
the final sediment vol ume tended to increase when the
initial settling rate decreased. A higher Fe2(SO4)3 con-
centration resulted in a higher packing efficiency for
pH > 4.0. At pH < 4.0, the final sediment vol ume
showed a maxi mum for a Fe2(SO4)3 concentration near
5 mN.
Since the suspensions were prepared with Na4PzOv,
the kaolinite particles were charged negatively because
the phosphate ions do not desorb readily. Hence, the
counterions according to DLVO theory are cations.
The results of the present study appear to be in qual-
itative agreement with DLVO theory (Pierre, 1992).
The connection with DLVO theory can be summarized
as follows.
"~ L6
- d /4 ~ ~ * pH = 4.0
.6. pH = 6.0
O pH = 9.3
1 2 A pH= 11.0
10' "  ' ' " ' ' ' ' ' ' ' ' " ' ' ' ' ' ' i , , n , - , - -
0 1 2 3 4 5 6 7 8 9 10
Fe2 ( SO4) 3 Concent r at i on ( mN)
Figure 6. Initial settling rate of 0.5% Na-kaolinite suspen-
sions as a function of Fe~(SO4)3 concentration, when floccu-
lation occurred.
When the Fe 3+ concentration was low, the electrical
double layer around the clay particles was thick and
the electrostatic repulsion between clay particles was
strong. The heaviest particles settled immediately un-
der gravity and formed the accumulated sediment at
the bottom of the cylinders. The colloidal particles
remained suspended for a longer period in a sol. Also,
because of the size distribution, the population density
of particles in this sol decreased progressively with an
increasing height in the cylinder. Hence, a diffuse tran-
sition-zone with the clear supernatant liquid was ob-
The addition of F& + electrolytes resulted in a re-
duction of the electrical double layer thickness. The
magnitude of the zeta potential is known to decrease
in these conditions (Young et al., 1987). When the
reduction in double layer thickness was moderate, lin-
ear aggregation of the clay particles was favored and
very open floes formed (Pierre, 1992). These flocsjoined
to each other and extended from wall to wall in the
test tube and formed the flocculated sediment, while
the supernatant liquid was free of particles. This floc-
culated sediment was able to flow. The very open floes,
as proved by the initial thickness of the sediment, are
responsible for a slow initial settling rate. However,
the fractal arms of the floes were flexible and could
slowly deform during aging in the final sediment. Con-
sequently, its vol ume decreased slowly with time.
With an increasing Fe 3§ electrolyte concentration,
the electrical double layer was even more compressed.
In this case, linear aggregation is known to become
progressively less important. Denser aggregation oc-
curred. The floes were less open, they settled faster,
but they were less prone to deformation during aging;
hence, a higher final sediment volume.
The mixed flocculation-accumulation behavi or oc-
curred when flocculation was very slow and in strong
competition with accumulation. Therefore, the final
sediment was composed of an accumulated layer and
a foccul at ed layer; the accumulated layer formed first,
Vol. 40, No. 5, 1992 Sedimentation of kaolinite 591
and t he flocculated layer after. Two sharp interfaces
coul d be followed successively in the same test tube:
(1) a sharp interface bet ween the accumul at ed layer
and t he remai ni ng suspension, at t he beginning; (2) a
sharp interface bet ween the flocculated layer and the
clear supernat ant liquid, at the end.
The effect of ani ons on the sedi ment at i on behavi or
in this clay can first be partly expl ai ned by the DLVO
theory, since the doubl e layer thickness decreases with
increasing val ence states of bot h the ani ons and the
cat i ons (Hi emenz, 1977). The fact that the flocculation-
sedi ment at i on occurred in a narrower range of con-
di t i ons with FeE13 electrolyte than with Fe2(SO4) 3 is in
agreement with the higher val ence state of SO42- com-
pared to C1 (Figure 3). They also had an effect on the
hydrolysis of Fe 3+.
Thi s study was not a study on the hydrolysis of Fe
electrolytes, whi ch was addressed in a large number of
previ ous publications. As ment i oned in the i nt roduc-
tion, it was known that fresh Fe electrolytes, as well as
fresh A1 electrolytes, favor the format i on of relatively
strong clay flocs whi ch do not collapse readily under
t hei r own weight (Swart zen-Al l en et aL, 1974). Our
purpose was si mpl y to det ermi ne a range of condi t i ons
where this happens, for further studies of the flocculate
structure, by supercritical drying, as with mont mor i l -
lonite (Zou and Pierre, 1992). Thi s structure study of
kaol i ni t e flocs is underway.
The rather abundant publications on Fe hydrolysis
and its action on clay can be summar i zed as follows
(Livage et al., 1988). Fe 3+ electrolytes experi ence their
own hydrolysis in aqueous medi a, and form their own
gels. When the electrolyte concentration increases, more
hydrolysis product s are formed. The Fe 3+ hydrolysis
product s i ncl ude a- FeOOH (geoethite) at pH > 9.0,
a-Fe203 (haematite) at pH < 4.0, and a mi xt ure of
bot h mat eri al s bet ween 4.0-9.0 (Livage et aL, 1988).
It is also known that the hydrolysis process and the gel
format i on process depend largely on the ani ons (NO3-,
C1 , SO42-). The hydrolysis of Fe2(SO4)3 is more com-
plex t han that of FeCI3. It gives rise to a larger variety
of compl exes such as Fe3(SO4)2(OH)5"2(H20),
Fe4(SO4)(OH)~o (Mat i j evi c et aL, 1975), or the insol-
uble compl ex Na[Al~304(OH)24(H20)~2(SO4)4] (Segal,
1984). In the presence of clay, these Fe 3+ hydrolysis
products exchange partly with Na +. They are also known
to adsorb on the surface of the clay particles, while
t hei r hydrolysis proceeds. Hence, they help to increase
the strength of clay flocs by the i nt ermedi at e of an Fe-
gel cement. Transmi ssi on electron mi crographs of the
Fe product deposi t ed on t he surface of clay particles
have been publ i shed (Bl ackmore, 1973; Rengasamy et
al., 1977; Greenl and et al., 1968; Oades, 1984).
The sedi ment at i on behavi or of fine kaolinite sus-
pensi ons in the presence of fresh Fe electrolytes can be
classified into three types: (1) sedi ment at i on by accu-
mul at i on of i ndi vi dual clay units; (2) sedi ment at i on by
settling of a flocculated clay whi ch forms a wall-to-wall
net work structure in the cylinder; (3) sedi ment at i on by
mi xed accumul at i on and settling of a flocculated layer,
when flocculation and accumul at i on are in strong com-
petition. Thi s occurs in condi t i ons i nt ermedi at e be-
t ween the 2 ot her kinds of sedi ment at i ons. The ob-
served behavi or, where the clay particles are charged
negatively, agrees at least qual i t at i vel y with the DLVO
Arora, H. S., and Coleman, N. T. (1979) The influence of
electrolyte concentration on flocculation of clay suspen-
sions: Soil Sci. 127, 134-139.
Blackmore, A.V. (1973) Aggregation of clay by the products
of iron(III) hydrolysis: Aust. J. Soil Res. 11, 75-82.
Bolland, M. D. A., Posner, A. M., and Quirk, J. P. (1976)
Surface charge on kaolinites in aqueous suspension: Aust.
J. Soil Res. 14, 197-216.
Flegmann, A. W., Goodwin, J. W., and Ottewill, R.H. (1969)
Rheological studies on kaolinite suspension: Proc. Br. Ce-
ram. Soc. 13, 31--45.
Georgia Kaolin Company, Inc. (1990) Information About
Properties of Hydrite UF Kaolinite Particles, Georgia Ka-
olinite Company, Union, New Jersey, 4.
Goldberg, S., and Glaubig, R.A. (1987) Effect of saturating
cation, pH, and aluminum and iron oxide on the flocculation
of kaolinite and montmorillonite: Clays & Clay Minerals
35, 220-227.
Greenland, D.J. (1975) Charge characteristics of some ka-
olinite-iron hydroxide complexes: Clay Miner. 10, 407-
Greenland, D. L., and Oades, J. M. (1968) Iron hydroxides
and clay surfaces: Trans. 9th Int. Cong. Soil Sci. Vol. I,
Hiemenz, P. C. (1977) Principles of Colloid and Surface
Chemistry: Marcel Dekker, Inc., New York, 352-452.
Livage, J., Henry, M., and Sanchez, C. (1988) Sol-gel chem-
istry of transition metal oxides: Prog. Solid State Chem.
18, 259-342.
Ma, K., and Pierre, A.C. (1992) Study on kaolinite flocculate
to be published.
Matijetivic, E., Sapieszko, R. S., and Melville, J. B. (1975)
Ferric hydrous oxide sols: I. Monodispersed basic iron(III)
sulfate particles: J. Colloid. Interface Sci. 50, 567-581.
Michaels, A. S., and Bolger, J. C. (1962) Settling rates and
sediment volumes of flocculated kaolin suspensions: I &
E.C. Fundam. 1, 24-33.
Oades, J. M. (1984) Interactionsofpolycationsofaluminum
and iron with clays: Clays & Clay Minerals 32, 49-57.
Pierre, A. C. (1992) The gelation of colloidal platelike par-
ticles: J. Can. Ceram. Soc. 61, 135-138.
Rand, B., and Melton, I. E. (1977) Particle interactions in
aqueous kaolinite suspensions; I. Effect of pH and electro-
lyte upon the mode of particle interaction in homoionic
sodium suspension: J. Colloid Interface Sci. 60, 308-320.
Rengasamy, P., and Oades, J. M. (1977) Interaction ofmo-
nomeric and polymeric species of metal ions with clay sur-
faces. I. Adsorption ofiron(III) species; II. Changes in sur-
face properties of clays after addition of iron(III): Aust. J.
Soil. Res. 15, 221-242.
Schofield, R. K., and Samson, H. R. (1954) Flocculation of
kaolinite due to the attraction on oppositely charged crystal
faces: Discuss. Faraday Soc. 18, 135-145.
592 Ma and Pierre Clays and Clay Minerals
Segal, D. L. (1984) Time dependent properties of colloidal
dispersions of ferric-hydroxy polycations: J. Chem. Tech.
Biotechnol. 34A, 355-362.
Swarten-Allen, S. E., and Matijevic, E. (1974) Surface and
colloidal chemistry of clays: Chem. Rev. 74, 385-400.
Swarten-Allen, S. E., and Matijevic, E. (1976) Colloid and
surface properties of clay suspensions, III. Stability ofmont-
morillonite and kaolinite: J. Colloid. Interface Sci. 56, 159-
van Olphen, H. (1977) An Introduction to Clay Colloid
Chemistry, 2nd ed: Wiley, New York, 89-105.
Worrall, W. E. (1986) Clays and Ceramic Raw Materials:
2nd ed., Elsevier Applied Science Publishers, New York,
Young, R. N., and Ohtsubo, M. (1987) Interparticle action
and rheology of kaolinite-amorphous iron hydroxide (fer-
rihydrite) complexes: App. Clay Sci. 2, 63-81.
Zou, J., and Pierre, A.C. (1992) SEM observations of"card-
house" structures in montmorillonite gels: J. Mater. Sci.
Lett. 11 10, 664-665.
(Received 27 March 1992; accepted 23 September 1992;
MS. 2219)